prajdabre commited on
Commit
ca36703
1 Parent(s): b0e8478

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +3 -2
app.py CHANGED
@@ -14,8 +14,9 @@ pad_id = tokenizer._convert_token_to_id_with_added_voc("<pad>")
14
  CREOLE = {"Hawaiian Pidgin": "hwc", "Saint Lucian Creole": "acf", "Belizean Creole": "bzj", "Chavacano Creole": "cbk", "Seychellois Creole": "crs", "Sranan Tongo": "srn", "Aukan": "djk", "Gullah": "gul", "San Andrés–Providencia Creole": "icr", "Jamaican Creole": "jam", "Mauritian Creole": "mfe", "Papiamento": "pap", "Pijin": "pis", "Tok Pisin": "tpi", "Torres Strait Creole": "tcs", "Australian Kriol": "rop", "Sango": "sag", "Saramaccan": "srm", "Bislama": "bis", "Nigerian": "pcm", "Sierra Leonean Creole": "kri", "Haitian Creole": "hat", "Kupang Malay": "mkn", "Tetun Dili": "tdt", "Malay Baba": "mbf", "Kituba": "ktu"}
15
 
16
 
17
- def generate(input, task, slang, tlang):
18
- lang = CREOLE[lang]
 
19
  inp = tokenizer(input.strip() + " </s> <2" + slang + ">",
20
  add_special_tokens=False, return_tensors="pt", padding=True).input_ids
21
  model_output = model.generate(inp, use_cache=True, num_beams=1, max_length=100, min_length=1, early_stopping=True, pad_token_id=pad_id,
 
14
  CREOLE = {"Hawaiian Pidgin": "hwc", "Saint Lucian Creole": "acf", "Belizean Creole": "bzj", "Chavacano Creole": "cbk", "Seychellois Creole": "crs", "Sranan Tongo": "srn", "Aukan": "djk", "Gullah": "gul", "San Andrés–Providencia Creole": "icr", "Jamaican Creole": "jam", "Mauritian Creole": "mfe", "Papiamento": "pap", "Pijin": "pis", "Tok Pisin": "tpi", "Torres Strait Creole": "tcs", "Australian Kriol": "rop", "Sango": "sag", "Saramaccan": "srm", "Bislama": "bis", "Nigerian": "pcm", "Sierra Leonean Creole": "kri", "Haitian Creole": "hat", "Kupang Malay": "mkn", "Tetun Dili": "tdt", "Malay Baba": "mbf", "Kituba": "ktu"}
15
 
16
 
17
+ def generate(input, slang, tlang):
18
+ slang = CREOLE[slang]
19
+ tlang = CREOLE[tlang]
20
  inp = tokenizer(input.strip() + " </s> <2" + slang + ">",
21
  add_special_tokens=False, return_tensors="pt", padding=True).input_ids
22
  model_output = model.generate(inp, use_cache=True, num_beams=1, max_length=100, min_length=1, early_stopping=True, pad_token_id=pad_id,